skills.rest
hidonguyen
hidonguyenCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

ck:team

Coordinates multiple Claude Code agent sessions for parallel research, implementation, review, and debugging.

Updated Apr 2, 2026
Search Tags:#debugging#orchestration#worktree#code-review#parallel-agents#team-management#agent-teams
One-click install
npx skills add https://github.com/hidonguyen/trade-ops --skill ck-team-hidonguyen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ck:team
Source: https://github.com/hidonguyen/trade-ops/tree/main/.opencode/skills/team
Command: npx skills add https://github.com/hidonguyen/trade-ops --skill ck-team-hidonguyen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill orchestrates multiple independent Claude Code sessions so a lead can run parallel research, implementation, review, and debugging workflows that require isolated contexts, direct inter-agent messaging, and shared task management, eliminating single-threaded bottlenecks and reducing coordination overhead.

Core Features & Use Cases

  • Templates for common workflows: research, cook (implement), review, and debug with defined step sequences and gating (plan approval, delegate mode).
  • Team lifecycle & tools: TeamCreate/TeamDelete, TaskCreate/TaskUpdate/TaskList, Agent spawning, and SendMessage for inter-agent coordination.
  • Worktree isolation for implementation: spawn developer teammates with git worktree isolation to avoid file conflicts and merge safely afterward.
  • Event-driven monitoring: TaskCompleted and TeammateIdle hooks reduce polling and provide reactive orchestration.
  • Agent memory & reporting: project-scoped memory, standardized report paths, and automatic synthesis of teammate outputs.
  • Use case: spawn three researchers to investigate architecture options, or spawn multiple devs and a tester to implement and validate a feature in parallel.

Quick Start

Use the skill by invoking a template with context such as: /ck:team research "payment reconciliation" --researchers 3 --delegate

Frequently Asked Questions about ck:team

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate parallel AI agents for code review and debugging?▼

Parallel AI agent coordination uses independent Claude Code sessions with isolated contexts and shared task lists to execute simultaneous code review and debugging workflows, eliminating single-threaded bottlenecks through direct inter-agent messaging and event-driven monitoring hooks.

What is the best way to run multiple AI agents in parallel for software development?▼

Running multiple AI agents in parallel is best achieved through orchestrated team workflows that spawn independent Claude Code sessions for research, implementation, review, and debugging, utilizing templates with defined step sequences and gating for plan approval.

Do I need git worktree isolation to run parallel AI agent teams?▼

Git worktree isolation is required specifically for implementation teammates to avoid file conflicts during parallel development. It allows multiple developer agents to work on isolated branches and merge safely afterward, though research and review workflows can operate without it.

Can I use parallel agent orchestration without Opus 4.6?▼

Parallel agent orchestration requires Opus 4.6 for all teammates within a CLI terminal environment with CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS enabled, alongside TeamCreate, Agent, and Task tool support to function correctly.

How does inter-agent messaging work in multi-agent code workflows?▼

Inter-agent messaging in multi-agent code workflows operates through SendMessage tools within Claude Code sessions, enabling direct communication between teammates for task delegation, status updates, and coordinated handoffs across isolated development contexts.

When should I not use parallel AI agent teams for development tasks?▼

Parallel AI agent teams should not be used for simple, sequential tasks that lack the complexity to benefit from isolated contexts, as the coordination overhead and requirement for Opus 4.6 with experimental flags outweigh the benefits of parallelization for straightforward workflows.

Related Skills

Explore More Software Engineering Skills→
auxi-wardrobeauxi-wardrobe
1

ck:team

Orchestrate multiple Codex sessions for parallel research, implementation, review, and debug workflows.

Official
Advanced
chen-001chen-001

worktree-status

Audit all git worktrees and report dirty state and merge status in a Markdown table.

Community
Intermediate
ramonclaudioramonclaudio
4

teams

Coordinate Claude Code sessions into a collaborative agent team with task graphs.

Community
Advanced
chrono-metachrono-meta
7

agent-composer

Map multi-part workflows to optimal AI agent combinations across Claude Code and Codex.

Community
Advanced